A confidential source indicated that while the framework remains tentative, the investment could exceed $10 billion. The Information originally broke the story on the developing alliance.

These talks follow an October reorganization at OpenAI that defined its Microsoft partnership boundaries, granting the startup greater latitude to court investors and collaborate across the AI ecosystem, according to CNBC.

Despite Microsoft's $13 billion support since 2019, an October release stated the software giant no longer possesses a right of first refusal for OpenAI's computing requirements. The report noted the firm may now co-develop products with third parties.

Amazon previously directed at least $8 billion toward Anthropic, yet CNBC suggests the retailer seeks further exposure to the generative AI sector. Microsoft similarly pledged up to $5 billion to Anthropic last month, joining Nvidia's $10 billion targeted investment.

AWS has engineered custom silicon since roughly 2015, providing essential hardware for model training. The provider introduced its Inferentia chips in 2018 and unveiled the newest Trainium processors earlier this month, the outlet reported.

Recent months saw OpenAI commit over $1.4 trillion to infrastructure, partnering with Broadcom, Advanced Micro Devices, and Nvidia. CNBC noted the company also finalized a $38 billion chatapacity deal with AWS last month, its inaugural pact with the cloud leader.

During October, a secondary offering totaling $6.6 billion allowed OpenAI staff to divest shares at a $500 billion valuation, the report concluded.

The post Could this Amazon deal be the biggest of them all? appeared first on www.israelhayom.com.

To substantiate its claim, the group released several files, photos, and documents allegedly verifying the breach under the title "Operation Octopus".

Following an inquiry by Israel Hayom, Bennett's office initially stated that he was "unaware of such an event." However, after political commentator Amit Segal reported that the phone had indeed been breached, Bennett's team rushed to issue a statement clarifying that the matter was being handled by relevant security and cyber officials and that the device in question is not currently in use.

Bennett did not stop there and issued yet another clarification, suggesting it is possible the phone was not breached at all. Amid the confusion, Israel Hayom contacted Shai Nahum, CEO of the tech firm CYGHT and a cyber warfare expert.

"According to forensic analysis of the leaked files, there is a high probability that this is not Bennett's phone, but apparently that of one of his associates," Nahum said.

He noted that the leaked photos and video clips had been processed by editing software known as Handbrake, "most likely to change the video format and edit various data points, including metadata."

Nahum added that an analysis of the leaked contact list indicates it does not belong to Bennett himself, but rather to someone likely close to him. For example, contacts are saved under names such as "Bennett's Minister," "Revital Bennett Comms," and even "Naftali Bennett."

"It appears there is an attempt here to portray Bennett as weak, as someone who does not know how to deal with cyber threats or analyze risks properly," Nahum emphasized. "It is understood that this is a group with political and regional interests operating under Iranian sponsorship."

"Sophistication level has risen"

Itay Schwartz, co-founder and CTO of the cyber company MIND, added: "Even before asking if Bennett's phone was truly hacked, one must understand what this event signals: The level of sophistication of cyberattacks in our region has risen dramatically. In recent years, we see actors like Hezbollah and Iran improving capabilities, after the Syrian Army's cyber unit also operated here for years. The enemy is simply improving.

"From the details published, it seems this is an old device that was not updated – exactly the type of weakness attackers know how to exploit. But it is important to emphasize: Even if there was a hack here, it does not mean the hackers received full access to all information. The moment an attacker enters a personal device, he can try to spread to additional assets, because everything is connected, but there is no certainty he succeeded," he added.

"The central lesson is that information security no longer stops at the device itself. A single civilian device today holds huge amounts of sensitive, personal, and sometimes classified information. To prevent such events, one needs not only to protect the phone – but to detect early when information starts moving to a place it shouldn't reach," Schwartz concluded.

The post Did Iranian hackers breach former Israeli PM's device? appeared first on www.israelhayom.com.

This revolution is rooted in civilian applications that have already become routine: drones that monitor agricultural fields and enable precise decision making; systems that inspect utility poles, bridges, and engineering facilities to detect faults in advance; tools that help rescue forces get an immediate operational picture; and more. Israel, with its combination of dense infrastructure and innovative technology industry, has become a natural testing ground where these applications continue to expand.

As civilian usage expands, so too does its significance. Drones conducting infrastructure patrols along power lines, detecting damage to pipelines, or surveying rooftops after a storm are no longer simple work tools. They operate in an environment where precision, reliability, and continuous operation are fundamental requirements. While these may be civilian applications, they are gradually becoming integrated into the fabric of daily life.

Expanding civilian applications into the realm of infrastructure also drives significantly more advanced developments. The Japanese Corporation NTT is developing an ultra-resistant drone capable of "absorbing" lightning strikes and channeling them to a safe spot on the ground, while researching ways to harness some of the energy generated by discharging. The fact that tasks that once required heavy, dedicated facilities can now be accomplished using a single drone or an array of drones highlights how this platform is expanding into roles that, until recently, were deemed impossible for this type of aircraft.

And when these abilities take shape, it's easy to envision where we're headed. While drones currently handle specific tasks, in the coming years they may operate entire infrastructure systems without human intervention. Systems that can monitor power lines, coordinate air traffic, and respond to malfunctions in real time. This is no longer a futuristic scenario but a natural extension of existing capabilities and ongoing development.

In this context, it is worth noting what is happening in Japan, which is advancing a vision of autonomous air mobility. The state is developing aerial corridors and traffic management systems for drones and eVTOLs, which makes it clear that low-altitude airspace is becoming a resource to be managed. This is a conceptual shift – drones no longer operate alone, but as part of a broad and coordinated array.

As civilian and infrastructural capabilities expand, the question is no longer what drones can do, but how they can be integrated into existing systems in an orderly fashion. As drones become components performing core tasks, they require coordination between entities, shared standards, and a uniform way to manage the skies in which they operate.

The challenge is not to slow down development but rather to build a conceptual framework and practical tools to allow the integration of these tools into vital infrastructure without creating congestion, risk, or lack of coordination. Rather than reacting once technology has already dictated reality, the gray area should be organized in advance, responsibilities clarified, thus ensuring that drones become a natural part of an operational system rather than an unwieldy addition.

Looking ahead, it is clear that drones are on their way to transition from mere technical objects to active components within a comprehensive infrastructure system. The question is not whether this will happen but how to make sure that skies filled with autonomous objects will become an effective, coordinated, and safe layer. A layer that supports our lives rather than happen to be alongside them when this transformation matures.

Asaf Green is the Director of Business Development at NTT Israel.

The post How drones are taking over the skies appeared first on www.israelhayom.com.

The meeting was organized by Amit Krig, senior vice president and head of Nvidia's development center in Israel. The emotional gathering at the company's US headquarters included the Israeli management team, who were at headquarters for professional meetings.

The post Nvidia founder meets former hostage employee appeared first on www.israelhayom.com.

Alongside Nvidia CEO Jensen Huang, the "Architects of AI" recognized by TIME include a cadre of rival and allied tech titans who are collectively reshaping the global landscape. Key figures featured in the report include Sam Altman of OpenAI, whose ChatGPT ignited the current boom; Mark Zuckerberg of Meta, who is integrating AI into social platforms; and Elon Musk, who is rapidly building massive data center infrastructure for xAI. The group also encompasses hardware and model leaders such as AMD CEO Lisa Su, Anthropic CEO Dario Amodei, and Google DeepMind CEO Demis Hassabis, as well as global investors and innovators like SoftBank's Masayoshi Son and Baidu's Robin Li, all of whom are described as "grabbing the wheel of history" to drive the technological revolution.

2025 was the year when artificial intelligence's full potential roared into view, and when it became clear that there will be no turning back.

For delivering the age of thinking machines, for wowing and worrying humanity, for transforming the present and transcending the… pic.twitter.com/mEIKRiZfLo

— TIME (@TIME) December 11, 2025

In the days leading up to the unveiling, reports proliferating on social media suggested the 2025 accolade belongs not to one person, but to a collective dubbed "The Architects of AI."

The a circulated cover art depicted a striking digital reimagining of the legendary "Lunch atop a Skyscraper" from 1932. Rather than steelworkers dining on a beam, the graphic showcases the tycoons and intellects engineering our digital destiny, according to Polymarket.

BREAKING: TIME Person of the Year reportedly leaked. pic.twitter.com/oe3okxsXoZ

— Polymarket (@Polymarket) December 10, 2025

The cover was confirmed to be authentic. Sitting center stage is Jensen Huang, the CEO of Nvidia, whose chips are powering this entire revolution, in no small part thanks to the Israeli company Mellanox he bought for a meager 6.9 billion dollars, which made the AI speed and bandwidth on his revolutionary chips possible to begin with. He's joined by household names like Elon Musk (Tesla, xAI), who made autonomous driving part of every day life, and Mark Zuckerberg (Meta), who transformed social media with AI. But the bench goes deeper, including OpenAI's Sam Altman, who made Chat-GPT a household name and revolutionized the way people retrieve data and interact with the world, and AMD's Lisa Su, whose chips have declared war on NVIDIA. Google DeepMind's Demis Hassabis is also on the leaked cover, a nod to the California giant that made AI latest breakthrough with Gemini 3.0, and made everyone question whether it is the next big thing – potentially outranking Chat-GPT because of its more sophisticated inner workings and easier supply chain. Also on the cover were another Chat-GPT rival, Anthropic's Dario Amodei, and the "Godmother of AI" herself, Fei-Fei Li, a Chinese-American computer scientist whose work on AI long before the revolution should be perhaps noted as the big bang of this new era.

It's a powerful visual: the people literally building the infrastructure of the 21st century, just like the steelworkers built the 20th.

The leak triggered chaos for gamblers. Viral posts on X noted that markets had "Artificial Intelligence" as the heavy favorite. However, when the cover showed people, the "AI" bet crashed. Since the winners are "The Architects of AI" (humans) and not "Artificial Intelligence" (the tech), the wager lost its value immediately.

When the cover leaked showing actual people instead of a generic robot, the betting rules kicked in. Since the title is reportedly "The Architects of AI" (humans) and not "Artificial Intelligence" (the tech), the value of the "AI" bet crashed from a sure thing to nearly zero in seconds.

Based on the betting odds chart from the various betting agencies, these were the following predictions:

Top political & world figures

• Pope Leo XIV (7–34%)

• Donald Trump (3–25%)

• Volodymyr Zelenskyy (11–12%)

• Jerome Powell (~9%)

• Benjamin Netanyahu (3–9%)

• Xi Jinping (~6%)

Cultural & media stars

• MrBeast (11–14%)

• Joe Rogan (8–12%)

The "tech" candidates (now likely part of the "architects")

• Jensen Huang (17–33%)

• Sam Altman (6–14%)

• Elon Musk (2–14%)

The post Time Person of the Year revealed amid leak chaos appeared first on www.israelhayom.com.

These processors – ranked second in power within Nvidia's inventory – are critical components for artificial intelligence workloads. However, the president clarified that the agreement does not cover Nvidia's highly desired Blackwell series or the next-generation Rubin units.

"I have informed President Xi, of China, that the United States will allow NVIDIA to ship its H200 products to approved customers in China, and other countries, under conditions that allow for continued strong National Security. President Xi responded positively!" Trump wrote.

Nvidia, currently the most valuable corporation globally, has surged due to booming AI demand, yet this success has positioned it directly within geopolitical tensions over technological supremacy.

Trump's message included the stipulation: "25% will be paid to the United States of America," seemingly indicating the federal government will claim a quarter of the revenue Nvidia secures from these sales, according to CNN.

The announcement follows a meeting between Trump and Nvidia CEO Jensen Huang, amidst the president's repeated assertions that America must secure victory in the global AI race against heavy Chinese investment.

"We applaud President Trump's decision to allow America's chip industry to compete to support high paying jobs and manufacturing in America. Offering H200 to approved commercial customers, vetted by the Department of Commerce, strikes a thoughtful balance that is great for America," an Nvidia representative told CNN. Additionally, Trump revealed that the Commerce Department is finalizing similar protocols for Intel, AMD, and other American manufacturers.

"AMD commends President Trump's leadership in finding a thoughtful approach to export policies that allows for US companies to compete globally," an AMD spokesperson stated. "The president's decision strengthens American competitiveness, supports high-value domestic jobs, and drives American investment and growth in the semiconductor industry."

CNN noted that Intel did not immediately reply to a request for comment. This authorization follows earlier agreements Trump established this year, including a deal for AMD and Nvidia to remit 15% of their China-sales revenue to the US government, while he also stated in August that the state plans to take a 10% equity position in Intel.

The post US to allow Nvidia H200 sales to China under new deal appeared first on www.israelhayom.com.

Like so many stories worth telling these days, Rosenberg's also begins in the early morning hours of that cursed Saturday. "Right after I understood the dimensions of the event, I realized it would also affect our employees," he said. "Nvidia has more than 5,000 employees in Israel, many of whom live and work in the south, and we started trying to find out what was happening with them."

"We were always there with them in the evening, from seven to nine, in sirens and missiles, and all this time we also had to take care of other employees who were affected, evacuated from home, or who lost family members and friends," he said. "It was a difficult and somewhat crazy routine. Nvidia had five employees whose family members were kidnapped to Gaza, and some were murdered there, and we even recruited new employees over time who also have relatives who are hostages. In addition, employees lost brothers and friends in the war, and company employee Amit Chayut fell in battle. We understood that we had to do everything we could to help them. It was clear to me that this was my most important mission and that of the company."

These years have been dramatic not only for Nvidia itself, but for the entire tech world. "I remember the pivotal day when they announced OpenAI's ChatGPT, exactly three years ago," Rosenberg said. "We all at the company understood that something had fallen, that the AI field was really breaking through, as we hoped would happen. This, of course, didn't happen in one day. The company built itself toward this moment. OpenAI worked with Mellanox on the technology back then, and we knew how important our products were for the AI era. It was a gradual rise until explosion – and when it happened, it happened fast and in big leaps, and it was important to preserve the organization's identity."

"Satisfaction is not just good conditions," Rosenberg explained. "Money is, of course, important and critical – salary, stocks, financial horizon – but a leading company needs to create satisfaction and interest and security for employees. From the CEO to the most junior manager, it's clear to everyone that we need to ensure our employees have it good and that they will want to continue working with us – and we invest a lot in this."

"On October 7, I didn't have a 'playbook' that says what to do if a company employee is kidnapped. We didn't know how to deal with something like this, but I hope I did the right things along the way. I also have no idea if anything I did somehow helped bring Avinatan home, but I tried to give his family as much strength as possible, because they were in a very lonely and difficult place."

The post Why Nvidia exec became hostage families' voice appeared first on www.israelhayom.com.

The vulnerability was discovered by the Cato CTRL research group of cybersecurity company Cato Networks and originates in common AI tools, including Google's Gemini, Microsoft's Copilot, and Perplexity's Comet.

The research demonstrated primary attack scenarios in which attackers direct AI assistants to display fake phone numbers and links to users when they request customer service contact information for various organizations. The scenarios included extracting sensitive user data and sending it to malicious sources without the user's knowledge, stealing login credentials, displaying false information, and creating fake narratives that could influence the user and lead to wrong decisions.

The technique used by attackers is called HashJack. All they need to do is add a prompt – malicious instructions – to a legitimate website address and distribute it. Once a user loads the website address with the malicious addition in the browser, the instructions "communicate" with smart AI assistants, such as Google's Gemini or Microsoft's Copilot, and trigger attack scenarios.

According to Cato Networks, traditional defense systems do not detect the attack because they operate through prompts (instructions) embedded in the website address after the hashtag symbol # in a process that does not leave the browser's work.

The attack exploits users' trust in legitimate websites by using link addresses that appear legitimate. The user has no reason to suspect at any stage of the process, unlike phishing sites that look suspicious. This way, any legitimate site could become an attack tool – with attackers not even needing to breach the site itself. They exploit how AI browsers interpret instructions after the hashtag symbol. This effectively creates a new subcategory of cyber threats in the AI world.

According to the company's statement, the companies whose tools the vulnerabilities were identified in were informed well in advance of the problems so they could address them before users were exposed to threats (a practice known in the cyber field as "white hat hacker" hacking). According to Cato's data, a fix was applied in the Copilot for the Edge browser on October 27, 2025. In the Comet browser, a fix was reported to have been applied on November 18, 2025. In the Gemini for Chrome browser, as of November 25, 2025, the problem has not yet been resolved.

The post Israeli tech firm finds AI vulnerability – Gemini susceptible appeared first on www.israelhayom.com.

A CNBC report that key customer Meta could partner with Google to use its Tensor Processing Units (TPUs) caused a 3% dip in Nvidia's stock. Nvidia asserted its chips are more flexible and powerful than the custom ASIC chips like Google's TPUs, claiming its product "offers greater performance, versatility, and fungibility than ASICs," CNBC reported.

We're delighted by Google's success — they've made great advances in AI and we continue to supply to Google.

NVIDIA is a generation ahead of the industry — it's the only platform that runs every AI model and does it everywhere computing is done.

NVIDIA offers greater…

— NVIDIA Newsroom (@nvidianewsroom) November 25, 2025

Analysts report Nvidia holds more than 90% of the AI chip market, but Google's proprietary chips are gaining attention as a potent, less costly substitute. Unlike Nvidia, Google does not sell its TPUs directly but offers them for lease through Google Cloud and uses them internally.

Earlier this month, Google launched its highly-rated Gemini 3 AI model, which was trained on its custom TPUs, not Nvidia GPUs. A Google spokesperson commented that demand for both custom TPUs and Nvidia GPUs is accelerating, and Google "We are committed to supporting both, as we have for years."

Nvidia CEO Jensen Huang addressed the TPU competition in his recent earnings call with investors, noting Google buys his GPUs and that Gemini can run on Nvidia technology. Huang confirmed contact with Demis Hassabis, Google DeepMind CEO, who texted that the theory on creating stronger AI by using more data and chips – known as "scaling laws" – is "intact." Nvidia believes this will drive even more demand for its systems, CNBC reported.

Hassabis texted that the theory on creating stronger AI by using more data and chips – known as "scaling laws" – is "intact." Nvidia believes this will drive even more demand for its systems, CNBC reported.

The post Nvidia vs. Google: Chip war escalates appeared first on www.israelhayom.com.

Amid all the upheaval there is one profession that many people overlook. Perhaps it is because it lacks glamour, or because it is mistakenly viewed as something simple to replace. The field of software testing remains one of the most stable areas in the industry, almost against all expectations.

One of the recurring misunderstandings in discussions about job replacement is the assumption that once a machine can perform a task it also understands it. AI can learn only from what has already happened. It identifies patterns, reconstructs situations and predicts the next step based on existing data. What it cannot do is sense that something feels wrong, that something is not functioning properly even when everything appears to be in order.

Software testers operate exactly in this space of things that have not yet occurred. They approach a system not as a chain of commands but as something that can surprise, break or behave in an unexpected way. They rely on intuition and human reasoning. Just as an experienced chef brings a personal touch that cannot be taught, professional QA testers bring their own kind of insight that sets them apart.

Anyone who has tried to replace testers in the past knows that promises of full automation rarely hold up. There were predictions that automation would eliminate the need for human testers. Reality proved the opposite. Automation tools were created but more people were needed to understand the system, design test plans, maintain the tools and catch what automated processes inevitably missed.

Professionals who work closely with AI driven systems know that they often require even more complex testing. Their outputs can be inconsistent; there is rarely a single correct answer and even a minor adjustment to a model can completely change how it behaves. Trusting AI to evaluate another AI system is an almost certain path to losing control. Human judgment remains indispensable.

The clearest examples come from actual field experience. Testers uncover bugs that automated tools would never identify. AI transcription services sometimes produce three different responses to the same query. Models miss basic business logic. Systems react unpredictably when the network is unstable, when a user double clicks or moves too quickly between screens. In every case it is a human who catches the issue in time, not an algorithm.

And so, while many other fields are experiencing upheaval, QA is actually gaining strength. The more technology the world invents, the more code we write, and the more AI we integrate into existing products, the greater the need for human eyes to examine everything and ensure that what we build truly works, is genuinely stable, and is fully ready for the people who operate these systems.

Marsel Vaida is the CEO and founder of Good Quality, a company specializing in software testing and automation development, based in Ofakim, with operations in the United States and South America.

The post AI surges, yet QA testing becomes more essential than ever appeared first on www.israelhayom.com.